Invoke-ScriptAnalyzer: An item with the same key has already been added. What is the proper syntax?

[RylandDeGregory]

I seem to be receiving a rule-less error whenever running PSScriptAnalyzer with default settings against a PowerShell script containing the following code block.

Environment

MacOS 10.15.4
PowerShell 7 LTS for MacOS
PSScriptAnalyzer 1.19.0

Code

$result = Invoke-Command -ComputerName $Computername -Credential $RemotingCred -Authentication Negotiate -ScriptBlock {
    $output=@()
    foreach ($item in $list) {
        $out = [PSCustomObject]@{
                    Property1   = $item.Property1
                    Property2   = $item.Property2
               }
        $output += $out
    }
    return $output
}

Command

Invoke-ScriptAnalyzer -path ./test.ps1

Expected Output

none

Actual Output

Invoke-ScriptAnalyzer: An item with the same key has already been added. Key: output

Question

What is the proper PowerShell syntax to prevent this error from generating? Or, which Rule can I ignore to suppress it? The code executes fine, just for the record :)

Thanks!

[bergmeister]

Can repro with PSSA 1.19.0 on PS 7 on Windows. Also happens with the current version of PSSA on master.

[rjmholt]

Would you be able to collect the full error print out with Get-Error and posting it here -- in particular the stack trace is important

[bergmeister]

I've got it in the debugger already, it's happening here:

PSScriptAnalyzer/Rules/UseUsingScopeModifierInNewRunspaces.cs

Line 220 in f9661f9

variableDictionary.Add(variableName);

Fix should be trivial, .Net 4.5.2 doesn't contain the .TryAdd method unfortuantely.

if (!variableDictionary.ContainsKey(variableName))
{
    variableDictionary.Add(variableName, variable);
}

I've opened a draft PR for it already, just need to write a regression test.

[bergmeister]

Actually, my fork didn't have the last 3 commits, this issue was already fixed in PR #1493 and is effectively a duplicate of #1492

[jantari]

Hello.

This critical bug is causing all my GitHub Actions CI that use PSScriptAnalyzer to hang indefinitely.
See screenshot:

And then it never does anything - doesn't quit either. I've had these running for over 45 minutes.

Using the ubuntu-latest runner with fresh PSSA installs from PSGallery every time.

Since this is such a severe bug, when can we expect a point release with the fix?

Thanks

[RylandDeGregory]

Agreed.

[bergmeister]

@jantari A hang is unexpected and might be a different issue, can you provide more details please?
For the moment, we suggest to pin the version of PSScriptAnalyzer using the -MaximumVersion 1.18.3 parameter on Install-Module?
What do you think about a 1.19.1 release @JamesWTruher @rjmholt ? If, so then I'd like to have PR #1497 and optionally #1491 considered as well.

[DigitalNightmare]

I am also seeing this issue, it seems to only affect scripts with hash tables that have duplicates in them. I've pinned to version 1.18.3 and get the error, which breaks my pester tests. like above. I also have tried pinning to 1.91.1 this throws the error still but the pester step in my build finishes without error, sans any testing it would otherwise do....

[rjmholt]

I'm in favour of a 1.19.1 release

[thomasrayner]

Hey, @rjmholt, is there a schedule for a 1.19.1 release, by any chance?

[rjmholt]

Best to ask @JamesWTruher and @SydneyhSmith

[SydneyhSmith]

@thomasrayner currently we are targeting a 1.19.1 release in about two weeks (likely the last week in July) however this is truly not a promise/commitment--just a good faith estimate....we are hoping to get another bug fix into this release, and then prioritize the release amongst other commitments

[jantari]

I wanted to say that during some general CI maintenance I switched to using 1.19.1/latest again and my deadlock is not occuring anymore, so appears fixed.

As far as I'm concerned, this can be closed.

EDIT: nvm it was closed all along